Run#

Before being able to run JADE, the instructions contained in Installation and Configure JADE must be followed.

Configure the run#

Once the software is installed and properly configured the user can select through the GUI which cases to run.

To open the run configuration GUI the user can move to the JADE root directory and run from the command line:

jade --rungui

python -m jade --rungui

A window like the one below should open:

../_images/rungui_bench.PNG — Benchmark selection TAB#

In this tab the user can select wich benchmarks to run and with which transport codes. The following is a brief description of the different columns contents/option:

Benchmark Name: Name of the benchmark. This is the string used in all folders, config files and input files related to that benchmark.folder containing input files for all codes.
Description: This is the extended name of the benchmark, this name will appear in specific outputs of the post-processing.
OnlyInput: When this field is set to True the benchmark input is only generated but not run. This can be useful when the user wants to run the benchmark on a different hardware with respect to the one where JADE is being used. The input will be generated for each selected code.

See also

External Run of a benchmark
MCNP: Runs (or only generate) the benchmark input for MCNP.
Serpent: Runs (or only generate) the benchmark input for Serpent.
OpenMC: Runs (or only generate) the benchmark input for OpenMC.
D1S: Runs (or only generate) the benchmark input for D1SUNED.
ACTINV: Runs (or only generates) the deterministic activation input for ACTINV. The NPS cut-off is unused for this code.
NPS cut-off: Number of histories to be simulated. Default values should provide sufficient statistical convergence but users are free to increase them.
Custom input: This columns allows to provide custom inputs to the different benchmarks. For the moment, this is used only in the Sphere Leakage and Sphere SDDR benchmarks where, if a number n is specified, this will limit the test to the first n isotope and material simulations (useful for testing).

Each benchmark that has at least one code (or the Only Input flag) selected will appear in the “Libraries” tab:

../_images/lib_tab.PNG — Libraries selection TAB#

Users need to select at least one library to be used for each benchmark or the run will not be performed for that benchmark.

Once also the libraries are selected the user can save the settings using the “Save Settings” button.

Important

The user is free to save the file to whatever location in case multiple settings need to be prepared. Nevertheless, JADE will always look for the <root>/cfg/run_cfg.yml file when the run is started. This is the file to be overridden and it is also the file that is loaded when the run GUI is opened. That is, settings of the previous run remain saved.

ACTINV inputs#

Select the benchmark and library in cfg/run_cfg.yml:

FNS-DecayHeat:
  codes:
    actinv: [TENDL2025-patched-ENDF8-JEFF33]
  description: FNS iron five-minute irradiation
  nps: 1
  only_input: false
  custom_input: null

nps is an unused compatibility field. ACTINV is deterministic, so increasing it does not improve the solution. The input must supply an inline 709-group spectrum and a positive-duration schedule. Mesh inputs, non-neutron projectiles, uncertainty sampling and other optional response calculations are outside this initial adapter. Configure the executable and nuclear data as described in ACTINV execution settings and ACTINV activation and decay data.

Run the benchmarks#

To run the selected benchmarks with the indicated transport code-library couples the user can simply run in the root directory from the command line:

jade --run

python -m jade --run

Continue executions#

In cases such as the Sphere Leakage benchmark, when running on HPC, it may preferable to run sequentially the different benchmark runs using a single job instead of submitting a different one for each benchmark. To handle this, JADE provides a continue run option that allows to identify all benchmarks that have been generated but not run successfully run them in sequence.

to resume simulations the user can run from the command line:

jade --cnt

python -m jade --cnt

For MCNP and D1S, the continue run will also automatically delete the .o output file so that the simulation can be restarted without errors.

Tip

The continue run is particularly useful when used in combination with the Only Input option.

ACTINV execution and continuation#

Each scalar ACTINV process has a 180-second timeout. Successful runs require valid finite output at every requested schedule endpoint before JADE writes actinv.complete. Once execution starts, a failed, timed-out or interrupted attempt removes any old completion marker. Unsupported execution settings are rejected before launching that case.

continue restarts an incomplete scalar case from its persisted input; it does not resume an internal solver checkpoint. Changing the selected data requires regenerating the input. See FNS decay heat: iron, five-minute irradiation for the initial case, data layout and availability.